This is the dockerfile with all ML stuff I use.
Guide how to install ML environment on Ubuntu
git clone https://github.com/illusionww/docker-jupyter-nn-tools
cd docker-jupyter-nn-tools
sudo chmod -R 777 .
sudo docker build -t illusionww/jupyter-nn-tools .
Your user should be added to the docker
group (sudo usermod -a -G docker $USER
). Alternatively, you can use sudo
to run the failed command as root. If you want to run container only for one GPU (if you have several) you can add NV_GPU=<GPU_NUM>
before command.
nvidia-docker run -ti --rm \
-v <YOUR HOST FOLDER FOR NOTEBOOKS>:/notebook -v <YOUR HOST FOLDER FOR DATA>:/data \
-p <YOUR EXTERNAL PORT FOR JUPYTER>:8888 -p <YOUR EXTERNAL PORT FOR TENSORBOARD>:6006 \
illusionww/jupyter-nn-tools
Example:
nvidia-docker run -ti --rm \
-v /home/illusionww/Documents:/notebook -v /data:/data \
-p 8888:8888 -p 6006:6006 \
illusionww/jupyter-nn-tools